• No results found

Appendix A: Data quality

Our profile of 300 children held in custody in the last six months of 2008 was based on data from the various documents contained on the secure access clearing house system (SACHS) held by the Youth Justice Board. The majority of information was collected from core Asset forms. Other documents used included pre-sentence reports (PSRs), post court reports (PCRs) and risk of serious harm forms (ROSH).

Asset is an assessment tool used by youth offending teams to collate information on all children who come into contact with the criminal justice system. It aims to identify a range of

background factors that may have some bearing on current and future offending behaviour, including family circumstances, education, physical and mental health as well as details of previous offending. Asset also contains information which can be used to assess the needs of the child, enabling appropriate interventions to be set up. Given the importance of Asset, we felt we should highlight some of the problems we encountered in collecting these data.

Sentenced sample

Extent of missing data

As is clear from our findings chapters, there were significant missing data across a range of variables. This included youth offending team (YOT) workers unable to determine whether a factor was relevant to the child (data was recorded as “don’t know”) and instances where information was incomplete. Examples are provided in the table below.

Different versions of Asset

The Asset was last amended in 2006 yet despite this, in 38% (N=75) of cases, data was collected from older versions of the Asset form, as these were the only forms available on SACHS relating to the offence/s in question. This meant there were some differences in the extent and nature of information available. For example, the old Asset form contained a different set of responses under ‘care history and ‘looked after’ status and had fewer questions on this issue compared to the new version.

Quality of completion

We encountered a number of problems related to the quality of the information on Asset. These included forms that were poorly filled out; for example, issues were highlighted in tick-box sections but no further details were provided in related open box sections. Whilst the Asset

Asset factor No. cases % cases

Accommodated by voluntary agreement 47 24%

Care order 53 27%

On child protection register 58 29%

Special educational needs 45 23%

No fixed abode 19 10%

form is expected to be completed at the beginning, midway and end of sentence, we had instances where YOT workers had copied and pasted new updates onto old forms, resulting in tick box sections not being updated to reflect the new situation and confused and sometimes conflicting narratives of what has taken place. In some cases where a PSR was available, important information detailed in that report was missing from the Asset form. Our main objective was to collect data from the Asset form which was linked to the offence recorded in SACHS for which the young person had received a custodial sentence. However, where that was incomplete or was of poor quality, we also reviewed earlier forms to collect background variables. Table A2 provides a breakdown of the type of Asset that was reviewed.

ROSH flagged but not available

As part of the Asset form, an assessment is made of whether a young person is at risk of causing harm to others and if this is a concern, a separate ROSH form is completed. There were 123 children classified as being a ROSH, but for 25 children (20%), no additional ROSH form was available on SACHS. For 17 children, ROSH was not highlighted in the Asset but a form was available, possibly in reference to a previous offence.

Remand sample

Many of the problems described in the sentenced section were also applicable to the remand sample (n=100). However there were a number of additional constraints for data collection. For example, a larger proportion of the sample (n=24) had a bail Asset, which covered the same substantive areas as the core Asset but were not as detailed (see Table A3). Only 27 of the sample had a core Asset available which was linked to the offence for which they were being remanded. A further 31 had core Assets for the appropriate timeframe (within a month of the court date) but the offences on SACHS did not match with those outlined in the Asset.

Table A3: Data collection on the remand sample

Asset validity N

Core Asset – offence match and completed one month prior or after sentencing 147 Core Asset – offences do not match but completed one month prior or after sentencing 29 Core Asset – offences match and completed within 3 months of sentencing 6

Core Asset – offences do not match and completed within 3 months of sentencing 9 Bail Asset – offence match and completed one month prior or after sentencing 3

Core Asset – over three months old 6

Asset validity N

Core Asset – offence match and completed one month prior or after court date 27 Core Asset – offences do not match but completed one month prior or after court date 31

Core Asset – over 3 months old 13

Bail Asset – offence match and completed one month prior or after sentencing 20

Bail Asset – no date, offences do not match, illegible 4